Introduction: Cancer screening programs are pivotal in facilitating early tumor detection and reducing cancer-related mortality. However, the reliance on individual performance indicators may not offer a fully integrated perspective of screening effectiveness. This study aimed to develop a composite indicator that consolidates key screening metrics using Principal Component Analysis (PCA) to derive objective weights. Methods: Data were drawn from the annual collection of breast and colorectal cancer screening activities conducted in 2023 by the eight regional health authorities (ATSs) of Lombardy, Italy. Indicators spanning structural, organizational, clinical-diagnostic, and impact dimensions were standardized and analyzed using PCA. The resulting factor loadings informed weights for computing a weighted arithmetic mean of the z-scored indicators. Robustness was evaluated through sensitivity analyses involving alternative normalization techniques. Additionally, comparisons were made against an equally weighted arithmetic mean of the same indicators. Results: The composite indicator identified marked geographical variations in screening performance across Lombardy’s ATSs. Structural and organizational factors, particularly examination coverage and participation rates, exerted the greatest influence on overall effectiveness in both breast and colorectal screening programs. In contrast, clinical-diagnostic and impact measures, as well as invitation coverage, contributed less substantially. High correlations in the sensitivity analyses affirmed the reliability of the proposed methodology. Conclusions: Adopting a composite indicator can substantially improve the evaluation of cancer screening initiatives by integrating diverse metrics into a cohesive, objective framework. In doing so, it enables more effective resource allocation, enhances adherence strategies, and guides targeted interventions aimed at bolstering screening success.
Improving Cancer Screening Program Evaluation Through Principal Component Analysis-Driven Composite Indicators: Evidence From Lombardy, Italy
Murtas R.Primo
Supervision
;
2025-01-01
Abstract
Introduction: Cancer screening programs are pivotal in facilitating early tumor detection and reducing cancer-related mortality. However, the reliance on individual performance indicators may not offer a fully integrated perspective of screening effectiveness. This study aimed to develop a composite indicator that consolidates key screening metrics using Principal Component Analysis (PCA) to derive objective weights. Methods: Data were drawn from the annual collection of breast and colorectal cancer screening activities conducted in 2023 by the eight regional health authorities (ATSs) of Lombardy, Italy. Indicators spanning structural, organizational, clinical-diagnostic, and impact dimensions were standardized and analyzed using PCA. The resulting factor loadings informed weights for computing a weighted arithmetic mean of the z-scored indicators. Robustness was evaluated through sensitivity analyses involving alternative normalization techniques. Additionally, comparisons were made against an equally weighted arithmetic mean of the same indicators. Results: The composite indicator identified marked geographical variations in screening performance across Lombardy’s ATSs. Structural and organizational factors, particularly examination coverage and participation rates, exerted the greatest influence on overall effectiveness in both breast and colorectal screening programs. In contrast, clinical-diagnostic and impact measures, as well as invitation coverage, contributed less substantially. High correlations in the sensitivity analyses affirmed the reliability of the proposed methodology. Conclusions: Adopting a composite indicator can substantially improve the evaluation of cancer screening initiatives by integrating diverse metrics into a cohesive, objective framework. In doing so, it enables more effective resource allocation, enhances adherence strategies, and guides targeted interventions aimed at bolstering screening success.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


